Many people are receiving unexpected sibling matches. Everyday on social media, “surprises” are being reported so often that they are no longer surprising – unless of course you’re the people directly involved and then it’s very personal, life-altering and you’re in shock. Staring at a computer screen in stunned disbelief.
Conversely, sometimes that surprise involves people we already know, love and believe to be full siblings – but autosomal DNA testing casts doubt.
If your sibling doesn’t match at all, download your DNA files and upload to another company to verify. This step can be done quickly.
Often people will retest, from scratch, with another company just for the peace of mind of confirming that a sample didn’t get swapped. If a sample was swapped, then another unknown person will match you at the sibling level, because they would be the one with your sibling’s kit. It’s extremely rare, but it has happened.
If the two siblings aren’t biologically related at all, we need to consider that one or both might have been adopted, but if the siblings do match but are predicted as half siblings, the cold fingers of panic wrap themselves around your heart because the ramifications are immediately obvious.
Your full sibling might not be your full sibling. But how can you tell? For sure? Especially when minutes seem like an eternity and your thoughts are riveted on finding the answer.
This article focuses on two tools to resolve the question of half versus full siblingship, plus a third safeguard.
Half Siblings Versus Step-Siblings
For purposes of clarification, a half sibling is a sibling you share only one parent with, while a step-sibling is your step-parent’s child from a relationship with someone other than your parent. Your step-parent marries your parent but is not your parent. You are not genetically related to your step-siblings unless your parent is related to your step-parent.
Ideally two people who would like to know if they are full or half siblings would have both parents, or both “assumed” parents to compare their results with. However, life is seldom ideal and parents aren’t always available. Not to mention that parents in a situation where there was some doubt might be reluctant to test.
Furthermore, you may elect NOT to have your parents test if your test with your sibling casts doubt on the biological connections within your family. Think long and hard before exposing family secrets that may devastate people and potentially destroy existing relationships. However, this article is about the science of confirming full versus half siblings, not the ethics of what to do with that information. Let your conscience be your guide, because there is no “undo” button.
Ranges Aren’t Perfect
The good news is that autosomal DNA testing gives us the ability to tell full from half-siblings by comparing the siblings to each other, without any parent’s involvement.
Before we have this discussion, let me be very clear that we are NOT talking about using these tools to attempt to discern a relationship between two more distant unknown people. This is only for people who know, or think they know or suspect themselves to be either full or half siblings.
Because the ranges of the amount of DNA found in people sharing close family relationships varies and can overlap. In other words, different degrees of relationships can be expected to share the same amounts of DNA. Furthermore, except for parents with whom you share exactly 50% of your autosomal DNA (except males don’t share their father’s X chromosome), there is no hard and fast amount of DNA that you share with any relative. It varies and sometimes rather dramatically.
The first few lines of this Relationship Chart, from the 2016 article Concepts – Relationship Predictions, shows both first and second degree relationships (far right column).
You can see that first degree relations can be parent/child, or full siblings. Second degree relationships can be half siblings, grandparents, aunt/uncle or niece/nephew.
Today’s article is not about how to discern an unknown relation with someone, but how to determine ONLY if two people are half or full siblings to each other. In other words, we’re only trying to discern between rows two and three, above.
As more data was submitted to Blaine Bettinger’s Shared cM Project, the ranges changed as we continued to learn. Blaine’s 2017 results were combined into a useful visual tool at DNAPainter, showing various relationships.
Note that in the 2017 version of the Shared cM Project, the high end of the half sibling range of 2312 overlaps with the low end of the full sibling range of 2209 – and that’s before we consider that the people involved might actually be statistical outliers. Outliers, by their very definition are rare, but they do occur. I have seen them, but not often. Blaine wrote about outliers here and here.
Full or Half Siblings?
So, how to we tell the difference, genetically, between full and half siblings?
There are two parts to this equation, plus an optional third safeguard:
- Total number of shared cM (centiMorgans)
- Fully Identical Regions (FIR) versus Half Identical Regions (HIR)
You can generally get a good idea just from the first part of the equation, but if there is any question, I prefer to download the results to GedMatch so I can confirm using the second part of the equation too.
The answer to this question is NOT something you want to be wrong about.
Total Number of Shared cM
Each child inherits half of each parent’s DNA, but not the same half. Therefore, full siblings will share approximately 50% of the same DNA, and half siblings will share approximately 25% when compared to each other.
You can see the differences on these charts where percentages are converted into cM (centiMorgans) and on the 2017 combined chart here.
I’ve summarized full and half siblings’ shared cMs of DNA from the 2017 chart, below.
||Average Shared cM
||Range of Shared cM
||1,317 – 2,312
||2,209 – 3,394
Fully Identical and Half Identical Regions
Part of the DNA that full siblings inherit will be the exact same DNA from Mom and Dad, meaning that the siblings will match at the same location on their DNA on both Mom’s strand of DNA and Dad’s strand of DNA. These sections are called Fully Identical Regions, or FIR.
Half siblings won’t fully match, except for very small slivers where the nucleotides just happen to be the same (identical by chance) and that will only be for very short segments.
Half siblings will match each other, but only one parent’s side, called Half Identical Regions or HIR.
Roughly, we expect to see about 25% of the DNA of full siblings be fully identical, which means roughly half of their shared DNA is inherited identically from both parents.
Understanding the Concept of Half Identical Versus Fully Identical
To help understand this concept, every person has two strands of DNA, one from each parent. Think of two sides of a street but with the same addresses on both sides. A segment can “live” from 100-150 Main Street, er, I mean chromosome 1 – but you can’t tell just from the address if it’s on Mom’s side of the street or Dad’s.
However, when you match other people, you’ll be able to differentiate which side is which based on family members from that line and who you match in common with your sibling. This an example of why it’s so important to have close family members test.
Any one segment on either strand being compared between between full siblings can:
- Not match at all, meaning the siblings inherited different DNA from both parents at this location
- Match on one strand but not the other, meaning the siblings inherited the same DNA from one parent, but different DNA from the other. (Half identical.)
- Match identically on both, meaning the siblings inherited exactly the same DNA in that location from both parents. (Fully identical.)
I created this chart to show this concept visually, reflecting the random “heads and tails” combination of DNA segments by comparing 4 sets of full siblings with one another.
This chart illustrates the concept of matching where siblings share:
- No DNA on this segment (red arrow for child 1 and 2, for example)
- Half identical regions (HIR) where siblings share the DNA from one parent OR the other (green arrow for child 1 and 2, for example, where the siblings share brown from mother)
- Fully identical regions (FIR) where they share the same segment from BOTH parents so their DNA matches exactly on both strands (black boxed regions)
If a region isn’t either half or fully identical, it means the siblings don’t match on that piece of DNA at all. That’s to be expected in roughly 50% of the time for full siblings, and 75% of the time for half siblings. That’s no problem, unless the siblings don’t match at all, and that’s entirely different, of course.
Let’s look at how the various vendors address half versus full siblings and what tools we have to determine which is which.
Ancestry predicts a relationship range and provides the amount of shared DNA, but offers no tools for customers to differentiate between half versus full siblings. Ancestry has no chromosome browser to facilitate viewing DNA matches but shared matches can sometimes be useful, especially if other close family members have tested.
Update 4-4-2019 – I was contacted by a colleague who works for an Ancestry company, who provided this information: Ancestry is using “Close Family” to designate avuncular, grandparent/grandchild and half-sibling relationships. If you see “Immediate Family “the relationship is a full sibling.
Customers are not able to view the results for ourselves, but according to my colleague, Ancestry is using FIRs and HIRs behind the scenes to make this designation. The Ancestry Matching White Paper is here, dating from 2016.
If Ancestry changes their current labeling in the future, this may not longer be exactly accurate. Hopefully new labeling would provide more clarity. The good news is that you can verify for yourself at GedMatch.
A big thank you to my colleague!
MyHeritage provides estimated relationships, a chromosome browser and the amount of shared DNA along with triangulation but no specific tool to determine whether another tester is a full or half sibling. One clue can be if one of the siblings has a proven second cousin or closer match that is absent for the other sibling, meaning the siblings and the second cousin (or closer) do not all match with each other.
Family Tree DNA
At Family Tree DNA, you can see the amount of shared DNA. They also they predict a relationship range, include a chromosome browser, in common matching and family phasing, also called bucketing which sorts your matches into maternal and paternal sides. They offer additional Y DNA testing which can be extremely useful for males.
If the two siblings in question are male, a Y DNA test will shed light on the question of whether or not they share the same father (unless the two fathers are half brothers or otherwise closely related on the direct paternal line).
FamilyTreeDNA provides Advanced Matching tools that facilitate combined matching between Y and autosomal DNA.
FamilyTreeDNA’s Family Finder maternal/paternal bucketing tool is helpful because full siblings should be assigned to “both” parents, shown in purple, not just one parent, assuming any third cousins or closer have tested on both sides, or at least on the side in question.
As you can see, on the test above, the tester matches her sister at a level that could be either a high half sibling match, or a low full sibling match. In this case, it’s a full sibling, not only because both parents tested and she matched, but because even before her parents tested, she was already bucketed to both sides based on cousins who had tested on both the maternal and paternal sides of the family.
GedMatch, an upload site, shows the amount of shared DNA as well. Select the One-to-One matching and the “Graph and Position” option, letting the rest of the settings default.
GedMatch doesn’t provide predicted relationship ranges as such, but instead estimates the number of generations to the most recent common ancestor – in this case, the parents.
However, GedMatch does offer an important feature through their chromosome browser that shows fully identical regions.
To illustrate, first, I’m showing two kits below that are known to be full siblings.
The green areas are FIR or Fully Identical Regions which are easy to spot because of the bright green coloring. Yellow indicate half identical matching regions and red means there is no match.
Please note that this legend varies slightly between the legacy GedMatch and GedMatch Genesis, but yellow, green, purple and red thankfully remain the same. The blue base indicates an entire region that matches, while the grey indicates an entire region not considered a match..
Fully identical green regions (FIR) above are easy to differentiate when compared with half siblings who share only half identical regions (HIR).
The second example, below, shows two half-siblings that share one parent.
As you can see, there are slivers of green where the nucleotides that both parents contributed to the respective children just happen to be the same for a very short distance on each chromosome. Compared to the full sibling chart, the green looks very different.
The half-sibling small green segments are fully identical by chance or by population, but not identical by descent which would mean the segments are identical because the individuals share both parents. These two people don’t share both parents.
The fully identical regions for full siblings are much more pronounced, in addition to full siblings generally sharing more total DNA.
GedMatch is the easiest and most useful site to work with for determining half versus full siblings by comparing HIR/FIR. I wrote instructions for downloading your DNA from each of the testing vendors at the links below:
Fraternal twins are the same as regular siblings. They share the same space for 9 months but are genetically siblings. Identical twins, on the other hand, are nearly impossible to tell apart genetically, and for all intents and purposes cannot be distinguished in this type of testing.
Here’s the same chart for identical twins.
23andMe also provides relationship estimates, along with the amount of shared DNA, a chromosome browser that includes triangulation (although they don’t call it that) and a tool to identify full versus half identical regions. 23andMe does not support trees, a critical tool for genealogists.
Unfortunately, 23andMe has become the “last” company that people use for genealogy. Most of their testers seem to be seeking health information today.
If you just happen to have already tested at 23andMe with your siblings, great, because you can use these tools. If you have not tested at 23andMe, simply upload your results from any vendor to GedMatch.
At 23andMe, under the Ancestry, then DNA Relatives tabs, click on your sibling’s match to view genetic information, assuming you both have opted into matching. If you don’t match your sibling, PLEASE be sure you BOTH have completely opted in for matching. I can’t tell you how many panic stricken siblings I’ve coached who weren’t both opted in to matching. If you’re experiencing difficulty, don’t panic. Simply download both people’s files to GedMatch for an easier comparison. You can find 23andMe download instructions here.
Scrolling down, you can see the options for both half and completely identical segments on your chromosomes as compared to your match. Above, my child matches me completely on half identical regions. This makes perfect sense, of course, because my father and my child’s father are not the same person and are not related.
Conversely, this next match is my identical twin whom I match completely identically on all segments.
Confession – I don’t have an identical twin. This is actually my V3 test compared with my V4 test, but these two tests are in essence identical twin tests.
The combination of these two tools, DNA matching and half versus fully identical regions generally provides a relatively conclusive answer as to whether two individuals are half or full siblings. Note the words generally and relatively.
There are circumstances that aren’t as clear cut, such as when the father of the second child is a brother or other close relative of the first child’s father – assuming that both children share the same mother. These people are sometimes called three quarters siblings or niblings.
In other situations, the parents are related, sometimes closely, complicating the genetics.
These cases tend to be quite messy and should be unraveled with the help of a professional. I recommend www.dnaadoption.com (free unknown parent search specialists) or Legacy Tree Genealogists (professional genealogists.)
The Final SafeGuard – Just in Case
A third check, should any doubt remain about full versus half siblings, would be to find a relative that is a second cousin or closer on the presumed mother’s side and one on the presumed father’s side, and compare autosomal results of both relatives to both siblings.
There has never been a documented case of second cousins or closer NOT matching each other. I’m unclear about second cousins once removed, or half second cousins, but about 10% of third cousins don’t match. To date, second cousins (or closer) who didn’t match, didn’t match because they weren’t really biological second cousins.
If the two children are full siblings meaning the biological children of both the presumed parents, both siblings will match the 2nd cousin or closer on the mother’s side AND the 2nd cousin or closer on the father’s side as well. If they are not full siblings, one will match only on the second cousin on the common parent’s side.
You can see in the example below that Child 1 and Child 2, full siblings, match both Hezekiah (green), a second cousin from the father’s side, as well as Susan (pink), a second cousin from the mother’s side.
If one of the two children only matches one cousin, and not the other, then the person who doesn’t match the cousin from the father’s side, for example, is not related to the father – although depending on the distance of the relationship, I would seek an additional cousin to test through a different child – just in case.
You can see in the example below that Child 2 matches both Hezekiah (green) and Susan (pink), but Child 1 only matches Susan (pink), from the mother’s side, meaning that Child 1 does not descend from John, so isn’t the child of the Presumed Father (green).
If neither child matches Hezekiah, that’s a different story. You need to consider the possibility of one of the following:
- Neither child is the child of the Presumed Father, and could potentially be fathered by different men
- A break occurred in the genetic line someplace between John and Hezekiah or between John and the Presumed Father.
In other words, the only way this safeguard works as a final check is if at least ONE of the children matches both presumed parents’ lines with a second cousin or closer.
And yes, these types of “biological lineage disruptions” do occur and much more frequently that first believed.
In the End
You may not need this safeguard check when the first and second methodologies, separately or together, are relatively conclusive. Sometimes these decisions about half versus full siblings incorporate non-genetic situational information, but be careful about tainting your scientific information with confirmation bias – meaning unintentionally skewing the information to produce the result that you might desperately want.
When I’m working with a question as emotionally loaded as trying to determine whether people are half or full siblings, I want every extra check and safeguard available – and you will too. I utilize every tool at my disposal so that I don’t inadvertently draw the wrong conclusion.
I want to make sure I’ve looked under every possible rock for evidence. I try to disprove as much as I try to prove. The question of full versus half siblingship is one of the most common topics of the Quick Consults that I offer. Even when people think they know the answer, it’s not uncommon to ask an expert to take a look to confirm. It’s a very emotional topic and sometimes we are just too close to the subject to be rational and objective.
Regardless of the genetic outcome, I hope that you’ll remember that your siblings are your siblings, your parents are your parents (genetic or otherwise) and love is love – regardless of biology. Please don’t lose the compassionate, human aspect of genealogy in the fervor of the hunt.
I receive a small contribution when you click on the link to one of the vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.
Thank you so much.